Background
The World Health Organization points out that 60% of the quality of an individual's health and life depends on his/her behavior and lifestyle [
1]. Health-risk behaviors are activities that increase a person's vulnerability or susceptibility to negative health outcomes [
2]. In contrast, health promoting behaviors entail a positive approach to living and a means of increasing well-being and self-actualization [
3]. Numerous publications have shown that practicing health promotion behaviors decreases the occurrence of disease and lowers the death rate [
4,
5]. A good health-promoting behavior depends on the living habits adopted during early years. Adolescents are at a dynamic transition period bridging childhood to adulthood, characterized by rapid, interrelated changes in body, mind and social relationships [
6]. There is a great deal of evidence that adolescents, particularly those in the age range 15-24, engage in health-risk behaviors such as smoking, drinking, having unprotected sexual intercourse, carrying weapons and adopting poor eating habits. These behaviors lead to a variety of adverse health outcomes, including major morbidities and mortalities among that age group [
7,
8], and will be carried into adulthood, jeopardizing their health status in later life [
9]. Many effects of health-risk factors among adults are avoidable if these behaviors are identified and changed at an early stage [
10,
11]. Therefore, it is essential to understand and evaluate health-promoting behaviors among adolescents in order to promote their healthy growth. In particular, this is an ideal time to conduct health education for adolescents in a higher education environment in terms of cost-effectiveness.
Many studies have shown that several factors at the individual, family and community levels influence an adolescent's risk for engaging in health-risk behaviors [
12‐
14]. However, there are fewer studies of adolescent health-promoting behaviors and particularly less information about the associated factors. A few studies have consistently shown a gender difference in health-promoting behaviors [
15]. Other factors that may be relevant include age, ethics, examination level at admission and educational level of their parents [
16‐
18].
To date, a few instruments have been developed to evaluate health-promoting lifestyles: Health-Promoting Lifestyle (HPLP) [
19], Adolescent Lifestyle Questionnaire (ALQ) [
20], Adolescent Lifestyle Profile (ALP) [
21] and Adolescent Health-Promoting (AHP) scale [
22]. The AHP scale based on the Pender's Health Promotion Model has been developed by Chen for evaluating adolescent health-promoting lifestyles. It has been widely used in Taiwan as a useful tool for investigating the health promoting behaviors of adolescents and evaluating the effects of health education [
23‐
29]. It has also been translated into many languages and its content has been examined cross-culturally [
30].
The development of higher education in mainland China has been very dramatic in recent years. For example, there were 5.56 million academy students and undergraduates in 2000. The number tripled (to 15.62 million) in 2005 and is expected to increase up to 20 million by 2010 [
31]. More and more students tend to practice health-risk behaviors such as smoking, alcohol drinking, sedentary lifestyle and irregular breakfasting [
32‐
35]. However, there have been few studies on health-promoting lifestyles among university students in mainland China. Moreover, no study so far has proposed or applied a standardized scale for examining such behaviors in this population.
Our aim in the present study is to examine the reliability and validity of the AHP scale for application to undergraduates in mainland China, investigating the prevalence of health-promoting behaviors and identifying the associated factors.
Methods
Subjects
The subjects were chosen from a total of 4523 undergraduates in Southern Medical University, located in Guangzhou, China. The university offers 22 specialties, stratified into nine medicine specialties and 13 non-medicine specialties. In this study, a two-stage stratified sampling method was applied. First, we applied a proportionate allocation strategy to sample two medicine specialties and three non-medical specialties, which were selected randomly on the basis of the discipline category-specialty sequence. Next, all the student numbers in the selected specialties were listed. A random sample of 100 student numbers was drawn from each selected specialty; 421 out of the total of 500 selected subjects agreed to be recruited for this investigation. The overall response rate was 84.2%. To evaluate the reliability of the results, 10% of the total respondents (44) were randomly selected for a retest 7 days after the baseline test.
Questionnaire
In view of the similarity in social and cultural, ethnic and demographic characteristics between Taiwan and mainland China, we adopted the AHP scale to assess the health-promoting lifestyles of university students. A detailed description of this scale can be found in previous publications [
22]. Briefly, it comprises 40 items assessing six dimensions of behavior: (1) nutrition; (2) social support; (3) life-appreciation; (4) health responsibility; (5) stress-management; and (6) exercise behavior. The frequency of reported behaviors was obtained using a self-reporting Likert scale with a five-point response format, "never, rarely, sometimes, usually, always", with the rating score ranging from 1 to 5.
In this study, the AHP scale was converted to a simplified Chinese version, then two linguists and Mei-Yen Chen, the developer of the scale, verified the accuracy of translation. As the AHP had not been used previously in mainland China, a pilot test was performed to determine whether it would be appropriate for our students. Thirty students completed the simplified Chinese AHP questionnaire for the purpose of testing the clarity and relevance of the statements. All items were understood perfectly and were completed without difficulty.
In addition, we collected some demographic information about each respondent: gender, grade, specialty, family residence, nationality, and monthly family income. The grade was divided into two categories: senior students (who had been in the university for 3~5 years) and junior students (who had been in the university for only 1~2 years). The family residence was classified into rural and urban areas on the basis of the local citizenship residence registered before enrollment.
Field work
Ethical approval was obtained prior to conducting this study. All respondents signed a written informed consent before participation, and each respondent was free to discontinue participation at any time. Because the study was intended to identify the usual pattern of university students' health practices and to avoid the confounding effects of seasonal holidays and the stressful period of examination time, the survey was conducted from November to December 2007, in the middle of the semester, via a self-administered questionnaire. To maximize the response rate without allowing the researchers to influence the respondents, all questionnaires were delivered and collected face-to-face by students trained as interviewers by the researchers. The respondents filled in the questionnaires by themselves. The interviewers on the site explained any unclear questions without inducement, if necessary. To examine the test-retest reliability of the questionnaire, about ten percent (44) of the respondents completed the questionnaires again a week later.
Data management
All valid questionnaires were input doubly into the database by two independent postgraduates using EpiData 3.1 software. The data were cleaned after double entry. We conducted computer logical checking and also manual checking, particularly when there were discrepancies between the two operators. In total, 14 invalid questionnaires were discarded because of incompleteness or double answers. Among these, nine had incomplete responses (i.e. more than 80% of the items were missing) and five had double answers to an item. Finally, a total of 407 valid questionnaires (96.9% of the total) were used for data analyses in this study. Of the 44 respondents who agreed to be re-interviewed, three questionnaires were rejected because they were not completed in line with the study protocol, resulting in 41 questionnaires for the retest analysis.
The missing values were handled using the method described in previous studies [
36]. In our survey, the item response rates were actually quite high. The average item response rates were 93.21% for the general information and 99.37% for the 40 items in AHP (range 98.45-99.91%). The raw score for each of the six AHP dimensions was derived by summing the item scores, and converted to a value for the dimension from 0 to 100. It was then re-calculated across the dimension as follows:
The AHP questionnaire was evaluated for reliability and validity. Split-half reliability was computed by correlating the scores of the odd half with those of the even half in each dimension of AHP. Test-retest reliability was assessed by the differences between test and retest scores using a paired-sample t test and was further assessed by the intraclass correlation coefficient (ICC). The internal consistency of the AHP items was assessed by Cronbach's
α coefficient. Construct validity was assessed by correlation analysis and factor analysis using principal component analysis [
37]. Student's t test and multiple linear regressions were applied to investigate the effects of individual characteristics on the health-promoting lifestyles of undergraduates. All these statistical methods were implemented via SPSS 17.0. The figures thus obtained were plotted using R project.
Discussion
Health-promoting lifestyle among adolescents has become a research focus worldwide. Life in college is a transitional period, offering good opportunities for establishing health-promoting lifestyles. Most research on health-promoting behaviors has been undertaken in the US and European countries, where university students are little engaged in health-promoting behaviors, especially healthy diet and physical activity [
39‐
41]. However, data on health-promoting lifestyles among university students in mainland China are limited. To our knowledge, this is the first (albeit preliminary) study to investigate health-promoting behaviors of university students in mainland China using a standardized scale. Chen et al. reported that the reliability and validity of the AHP scale was very satisfactory in studies of health-promoting behaviors in Taiwan [
22]. The present study evaluated the appropriateness of this scale for use with undergraduates, providing important information for a further large-scale investigation.
A Cronbach's
α value of 0.7 or higher is generally considered sufficient to demonstrate internal consistency [
42]. The Taiwan studies reported that the internal consistency of the AHP scale was higher than 0.7 on all six dimensions [
22]. Similarly, the present study found
α values above 0.7 for all dimensions except for Nutrition behavior, indicating that the internal consistency is satisfactory when the AHP scale is used in mainland China. The Nutrition behavior dimension had a relatively low Cronbach's
α coefficient (0.684), and the same applied to ICC and split-half reliability, indicating there might be some problems in the conceptualization of Nutrition behavior. It may be difficult for undergraduates to assess their daily intakes of various kinds of food.
Factor loadings larger than 0.4 are usually considered to support the factor construction of a particular dimension [
42]. Judged by this criterion, our factor analysis results indicated that the AHP scale accorded generally with the theoretical construction. Correlation analysis indicated that each of the 40 items was highly correlated with the hypothesized dimension, while relatively low correlations were observed between the items and other dimensions. In addition, consistent with previous studies, we found a significant gender difference in health-promoting behaviors, suggesting that this scale had good construct validity since it distinguished different subgroups as expected. Therefore, we may conclude that AHP was acceptable and applicable for evaluating the health-promoting lifestyles of adolescents in mainland China.
The results of this study showed that the mean scores of all dimensions of AHP were lower than 70. In particular, the mean scores on the Health-responsibility and Exercise behavior dimensions were lower than 60. This finding is consistent with the results of previous investigations in Taiwan [
43,
44] and Hong Kong [
15]. A sedentary lifestyle is a common and serious problem among university students. Compared to young adults in general, the pressure of work is so severe for university students that much of their time and energy is likely to be occupied with their studies. On the other hand, the popularization of computers and the Internet may provide more choices of entertainment and reduce interest in exercise. Lack of exercise facilities is also a major reason why university students do not participate actively in exercise.
We found that the health-responsibility score was the second lowest among the six dimensions. Undergraduates in mainland China usually reside on campus with schoolmates, and are less likely to pay attention to their own health than younger adolescents such as primary and high school students who live with parents and are frequently reminded about health. Moreover, university life adds more stress and requires more independent decision-making by young people. They are also challenged to attain the personal growth and perseverance necessary to cope with life stress and to establish healthy interpersonal relationships. All this is probably reflected in the finding that students considered themselves not to be doing well enough in Social support, Life-appreciation and Stress-management.
The food consumption patterns of university students are of particular concern because they also tend to skip meals frequently, eating "fast" foods and snacks. This may be understood because students eat in the school canteen where the service time is short and fixed, and food variety is limited in mainland China. For instance, fresh fruits are seldom available in the university canteen under investigation. As revealed by the survey of adolescent health risk behaviors in a Chinese city in 2005, the highest incidence of skipping breakfast occurred among the university student group [
45].
In addition, this study provides evidence of gender and grade differences in the choice of health-promoting lifestyles. Female students were more likely to take a healthy diet but males engaged in more physical exercise, which is consistent with the findings of the Hong Kong Federation of Youth Groups poll [
46]. Female students showed more confidence than male students in the Social support dimension (mean score 72.16 versus 63.54) and were far more capable than male students in the use of interpersonal relationships to maintain their psychosocial well-being.
Nutrition behavior was better among senior than junior students, probably because according to the curriculum, nutrition courses are provided to senior students. In terms of life-appreciation, this study revealed that junior students were far more capable than senior students, which may be because the senior students had less enthusiasm for university life owing to a longer time of sensitization.
Although personal characteristics such as gender affect health behaviors, they are seldom incorporated into health interventions since personal characteristics cannot be changed. Health education programs should be planned to stimulate the interests of different students according to their inclinations and characteristics.
There are at least four limitations in the present study. First, no detailed information about non-responders was collected. However, since the response rate was high (84.2%), the bias due to missing information on non-respondents, if any, should be small. Secondly, although the interviewers received uniform training, the interviewers' explanation might still have influenced the results, and this was difficult to evaluate. Thirdly, the subjects in this study were older adolescents with a mean age of 21 and a range of 16-25; the ages 15-25 are usually considered adolescent in mainland China. This should particularly be noted when our results are compared with findings from other regions such as the US, where 'adolescent' is defined as those under the age of 21. Lastly but most importantly, all the subjects were picked from only one university, so the results may provide useful information about only those students' health behaviors. It is not prudent to generalize the results to the whole population of university students or adolescents in mainland China. Students in a medical university are more likely to adopt a healthy lifestyle because of the influence of the medical environment, although this study showed no differences in any dimensions of AHP between medicine specialties and non-medicine specialties. Therefore, in this sense, a large-scale investigation at non-medicine universities should be launched for further study.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
DW conceived of the study, assisted with the survey, completed the statistical analyses and drafted the manuscript. CQO assisted with data analysis and paper drafting and finalized the manuscript. MYC participated in the design and coordination of the study. ND participated in literature retrieval. All authors read and approved the final manuscript.